HomeMy WebLinkAboutBoard of Selectmen Agenda April 27, 2011 Packet_201402061623560319 ________________________________________________________________________
447 Falls Bridge Road, Blue Hill, Maine 04614 phone: 207.374.3294 web: www.bluehillhydraulics.com
5 April 2011
Mr. Mohamed Nabulsi, Acting Director
Department of Public Works
188 Madaket Road
Nantucket, MA 02554
Dear Mr. Nabulsi:
Attached please find a draft report titled, “CFD Model Review - Children’s Beach
Stormwater Pump Station.”
This write-up contains a review of the Computational Fluid Dynamics (CFD) model used
by AECOM to investigate causes of operational problems at the pump station that
occurred in mid-October 2009.
Please contact me if you have any questions, or require additional information.
Sincerely,
John E. Richardson, Ph.D., P.E.
President
CFD Model Review - Children’s Beach Stormwater Pump Station
(Draft Report)
Submitted to:
Town of Nantucket
Contents
Page
1. Background ………………………………………………………… 2
2. Introduction ………………………………………………………… 2
3. CFD Model Review ……………………………………………....... 3
3.1 CFD Model Setup ….……………………...…...…………… 3
3.2 Study Scenarios ………………….…….…………………… 4
3.3 Critique ..…………..…………...…………………………… 5
3.4 Conclusions ………..…..……………………...……………. 8
3.5 Recommendations ………………………………...………... 9
4. Additional Analyses ……...………………….……………………... 10
4.1 CFD Model Setup …………………………………………... 10
4.2 Study Scenarios ……………………………………………. 12
4.3 Results ……………………………………………………… 14
5. Conclusions and Recommendations ………………………..……… 17
References ……………………………………………………………… 18
Appendix One: Cost Estimate for Installation of Variable Frequency Drives 19
CFD Model Review
Page 2
1. Background
Blue Hill Hydraulics was selected by the Town of Nantucket to provide peer review of the
Computational Fluid Dynamics (CFD) model used by AECOM to investigate causes of operational
problems at the Children’s Beach Pump Station that occurred in mid-October 2009.
As described in the request for proposals (RFP), the Town of Nantucket would like to verify that its
stormwater system in the area of Children’s Beach will perform satisfactorily under all hydraulic
conditions for the 2-year design storm with modifications recommended by AECOM. The suggested
modifications were based on the results of CFD modeling carried out by AECOM using a computer
software package known as FLOW-3D®.
2. Introduction
The Children’s Beach Pump Station was placed in operation during the second week of June 2009.
Throughout the summer, the pump station worked properly and no operational problems were reported.
However, in mid-October 2009 the pumps shut down during a storm event that occurred at high tide. To
investigate the cause of the problem, the town’s consultant (AECOM) constructed a CFD model of the
pump station and carried out analyses designed to identify the cause of the problem. As a result of
AECOM’s work, a number of possible failure scenarios were identified and suggestions for improving
the pump station’s design were made (AECOM [2010] and Parece [2010]).
In the summer of 2010, Blue Hill Hydraulics was asked to review the CFD modeling carried out by
AECOM. As part of this review, Blue Hill Hydraulics visited the project location in late August and
discussed operational problems with Mr. Mohamed Nabulsi (Town of Nantucket), Mr. Thomas Parece
(AECOM), and Mr. Jeffrey Willett (Town of Nantucket). As a result of the site visit, AECOM was
asked to supply the information listed in Tables 1 and 2 (Richardson, 2010). By the end of September
AECOM had provided all of the essential data required for the review with the exception of drawings
showing the layout of the pump station and downstream piping, and the FLOW-3D® output files. The
output files arrived in mid-October.
CFD Model Review
Page 3
Table 1: Required Data for CFD Model Review
Data Type Purpose
FLOW-3D input files (prepin.*) Review model physics
As-built drawings of pump station
and downstream piping
Verify model setup
Tide chart data for mid-October
storm
Check downstream boundary
conditions
Startup algorithm for pumps Identify startup water level for
pumps
Pump startup data (Q vs. t) Check upstream boundary conditions
FLOW-3D output files (flsgrf.*) Check computed results
Table 2: Additional Information
Data Type Purpose
Mid-October storm rainfall data Critique upstream boundary
conditions
Size of mid-October storm
(hydrologic characterization)
Critique upstream boundary
conditions
Results of any hydrologic studies
completed for upstream drainage area
Critique upstream boundary
conditions
Duck valve performance data (e.g.,
headloss characterization)
To estimate capacity of system
without pumps in operation
Failure mode(s) for pumps recorded
during mid-October storm
To compare with proposed
modifications
3. CFD Model Review (AECOM)
3.1 CFD Model Setup
AECOM’s CFD model includes the pump station, discharge pipe, and a short section of the pipe that
leads to the pump station from upstream. As constructed, the model is capable of simulating the flow of
water through the pump station (e.g., pump startup and shutdown). However, the model was not setup
to simulate the movement of air into and out of the pump station as a result of the water’s movement.
CFD Model Review
Page 4
3.2 Study Scenarios
As discussed in Parece (2010), AECOM identified three different operating scenarios that could lead to
pump failure. These are as follows:
(1) The first is when the influent stormwater pipe is completely submerged, as occurred
during the stormwater event in October 2009, and the pumps are called to turn on.
The pumps draw down the water level at such a rate that make-up air is unable to
enter the pump chamber quickly enough. This results in the pumps tripping out on
high electrical overload, since the pumps are drawing more power in order to
overcome the suction pressures created by the totally submerged influent pipe.
(2) The second scenario is when the influent stormwater pipe is completely submerged,
the pumps are running, and the pumps are called to turn off. As the pumps shut
down, the water column in the influent pipe continues to move towards the pump
chamber, creating an increase in pressure in the pump chamber for less than 1
second, which then quickly subsides.
(3) The third scenario involves the discharge side of the pump station structure. When
the stormwater pumps are called to turn-on and the discharge pipe is completely
submerged, water quickly fills the discharge chamber within a matter of seconds.
This situation causes a water hammer like phenomenon where the pressure in the
discharge chamber experiences a spike of pressure. The pressure within chamber
climbs to over 1,000 psi for less than a tenth of a second, then drops to about 250 psi
for about one-half of a second, before dropping to the normal operating pressure of
less than 6 psi. The increase in pressure occurs quickly and subsides as the water
column in the outfall pipe begins to move from a static condition. A similar
experience occurs during pump shutdown, although pressures are not as great
(Excerpted from Parece, 2010).
The first operating scenario was not simulated with the AECOM model. The last two scenarios were
simulated with the AECOM model and changes to the model were made so that improvements to the
pump station’s design could be investigated.
A critique of the study scenarios and modeling results is presented in the following section as well as a
critique of proposed modifications to the pump station.
CFD Model Review
Page 5
3.3 Critique
3.3.1 Study Scenarios
Three operating scenarios, that could lead to pump failure, were identified by AECOM. The first
concerned the availability make-up air during pump startup, the second concerned a water-hammer
effect on the inflow side of the pump station when pumps are shutdown, and the third concerned the
development of high-pressures on the discharge side of the pump station during pump startup.
3.3.1.1 Availability of Make-Up Air (Inflow Side)
Engineers in the field reported a loud whistling noise coming from a cover on the inflow side of the
pump station when the pumps were turned on. This led to speculation that make-up air (required to fill
the upstream side of the pump station when water levels are drawn down) cannot enter the pump
chamber quickly enough during certain startup conditions and could cause the pumps to draw
excessively high electrical loads.
The AECOM model was not used to simulate this operating condition. As noted earlier, the AECOM
model was setup to simulate water movement rather than the movement of air (or the movement of
water and air together).
At this time it is not certain whether or not this operating condition contributed to the failure of the
pumps in mid-October 2009. However, engineers did remove a cover on the inflow side of the pump
station and this change did not improve the operation of the pumps (personal communication with Mr.
Mohamed Nabulsi, Town of Nantucket).1
3.3.1.2 Pump Shutdown (Inflow Side)
Pump shutdown on the inflow side of the pump station was simulated with AECOM’s model.
According to Parece (2010), “as the pumps shut down, the water column in the influent pipe continues to
1 Ventilating the inflow side of the pump station is, however, recommended for the reason stated by Parece (reference – page
4, section 3.2, item [1]).
CFD Model Review
Page 6
move towards the pump chamber, creating an increase in pressure in the pump chamber for less than 1
second, which then quickly subsides.”
While discussed in the write-up, computed results for this scenario were not included in the 8 March
2010 submittal (Parece, 2010). Results for a similar scenario were, however, included and these results
are consistent with the comments above (see Section 5.2.1 for further discussion).
3.3.1.3 Pump Startup (Discharge Side)
Pump startup on the discharge side of the station was simulated with AECOM’s model for a condition
where the discharge pipe was full of water (presumably a condition where pump startup is coincident
with a high-tide). According to AECOM’s results this condition produces high pressures in the
discharge chamber for a brief period of time and could cause the pumps to shutdown.2
On 18 October 2009, the pumps at Children’s Beach failed in the early afternoon. According to
meteorological data supplied by AECOM – heavy rain on 18 October was reported at 12:53pm and a
high tide was predicted to occur at 12:44pm (3.9 ft). Given the generally good operational history of the
pump station, and the considerable length of the discharge pipe, it would seem that the heavy rain event
coincident with a high tide most likely caused the pumps to fail (i.e., for a condition such as this the
pumps would have to set a large body of water into motion over a very brief period of time).3
3.3.2 Proposed Design Changes
Except for a change to be made at the end of the discharge pipe, all of AECOM’s proposed design
modifications are based on the use of vents to reduce operating pressures in the pump station. This
approach makes sense; however, the basis for sizing the vents has not been provided.
2 No effort was made to relate pump performance (i.e., electrical demands) to the CFD modeling results. This, however,
would be difficult to do and does not diminish from the strength of the conclusion. 3 The pump station has operated properly at other periods of time when rains have been heavy. The difference here is that
the tide was high as well - filling the discharge pipe with water.
CFD Model Review
Page 7
3.3.2.1 Addition of Vents onto Roof of Pump Station (Modification 1)
The addition of six-inch vents on top of the pump station to allow make-up air to enter the chamber and
to reduce pressures when the pumps are turned-on and shutdown was recommended by AECOM
(Parece, 2010).
On the upstream side of the pump station ventilation should be added to reduce the amount of power
needed to start the pumps in the event that the influent pipe is completely full of water. Ventilating the
upstream side of the pump station would make it easier for make-up air to enter the chamber and it
would reduce the water hammer effect caused by the abrupt shutdown of the pumps. This latter
condition was identified by AECOM, but it is not clear whether or not it would be all that significant.
To more accurately estimate the size of these vents, a greater portion of the upstream piping should be
included in the model and some additional changes to the upstream boundary conditions should be made
to incorporate the dynamic characteristics of the drainage basin upstream.
Roof vents on the discharge side of the pump station are not mentioned in the write-up, but numerical
results that include these features appear in AECOM’s PowerPoint presentation (Parece, 2010).
Addition of these vents would certainly reduce pressures in the discharge side of the pump station, and
possibly keep the pumps from failing; however, the size of these vents would have to be carefully
considered. If the vents were not large enough, then pump failure would continue; and, if the vents were
too large stormwater would be pumped onto Children’s Beach.
The addition of roof vents as proposed by AECOM is sensible; however, additional work is required to
size them adequately. In addition to this, consideration must be given to whether or not it is acceptable
for stormwater “overflows” to be pump onto the beach (granted that if the vents were sized properly,
then the amount of this water be small compared to the overall discharge through the system).
3.3.2.2 Addition of Pressure Relief Valve (Modification 2)
The addition of a pressure relief valve in the separator wall between the inflow and discharge side of the
pump station was proposed to reduce pressures downstream of the pumps. No analysis work to support
this proposed modification was provided and its success would depend on the dynamics of flow within
CFD Model Review
Page 8
the pump station. For example, suppose that a high tide fills the discharge pipe and a rain event causes
the pumps to turn-on. The addition of the pressure relieve valve could cause stormwater to flow back
into the inflow side of the pump station and stormwater flowing into the station from upstream could
cause the station to fill entirely. The concern is whether or not pressures would continue to rise creating
– once again – problems for the pumps.
Precedence for the use of similar pressure relief valves in similar circumstances would be useful for
evaluating this proposed modification. A more rigorous basis for design would also help.
3.3.2.3 Addition of 90-Degree Elbow to Discharge Pipe (Modification 3)
The addition of a 90-degree elbow at the end of the pipe addresses concerns related to the discoloring of
boats in the marina where the stormwater is discharged. As described, the elbow would force
stormwater to be discharged below the water surface and “it is anticipated that [the addition of the
elbow] will minimize the amount of discoloring of the boats as the [flocculated iron] will settle out and
dissipate with the tides below sea level (Parece, 2010).”
Based on minimum gravity flows rates through the stormwater system and fact that the discharge flows
would be buoyant compared to the surrounding seawater – it is not certain that this proposed change
would work as intended. It is possible that, as proposed, the discharge water would float to surface and
not mix with the seawater and that the resulting flow patterns would not be very much different from
what occurs today.
3.4 Conclusions
The following conclusions were made as a result of this review.
(1) Pump failure in mid-October 2009 was most likely caused by a heavy rainfall coincident
with a high-tide (i.e., a condition such as this would cause the discharge pipe to fill with
water and the pumps would not be able to overcome the inertia of the water within the
discharge pipe when they were called upon to start up).
CFD Model Review
Page 9
(2) Without proper ventilation pressures within the pump station can be highly variable (as
shown by AECOM’s modeling); thus, one approach to reducing this variability and
improving the reliability of the pumps would be to add roof vents as proposed. To do this
correctly, the vents need to be sized properly and considerations for stormwater
overflows would have to be made.
(3) As described, the 90-degree elbow at the end of the discharge pipe will not necessarily
correct the coloration problem in the harbor.
3.5 Recommendations
As a result of this review it is recommended that:
(1) Additional analyses should be carried out to determine the benefit of increasing the start
up time for the pumps or reducing pumping rates to eliminate surcharging of the pump
station.4
(2) A mixing zone analysis should be carried out to confirm whether or not the addition of
a 90-degree elbow at the end of the discharge pipe would improve the dilution and
transport of flocculated iron into the harbor. The concept seems right, but questions
about available mixing energy (particularly for low “gravity” flows) remain.
(3) A set of drawings that accurately show the layout of what has been built should be
produced and used as reference material for future work. In addition to this, the
drawing should define the relationship between commonly used datums relevant to this
project.
4 NOTE: This additional work was carried out by Blue Hill Hydraulics at the request of the Town of Nantucket
and is presented in Section 4 of this report.
CFD Model Review
Page 10
4. Additional Analyses
4.1 CFD Model Setups
At the request of the Town of Nantucket, Blue Hill Hydraulics constructed its own CFD model of the
Children’s Beach Pump Station using information provided by AECOM (see Figure 1).
(a) (b)
(c) (d)
Figure 1: Children’s Beach Pump Station – Existing
(a) Pump Station, (b) Pump Station with Capping Removed viewed from Discharge Side,
(c) Pump Station with Capping Removed viewed from Inlet Side, (d) Cutaway
CFD Model Review
Page 11
The outlet pipe from the pump station was connected to a 331 ft long 48 inch diameter pipe as shown in
Figure 2(a) and the resulting model was used to simulate startup conditions on the discharge side of the
station – similar to what AECOM did in their analysis. In addition to this, the operation of a new outlet
chamber, that would discharge pumped flows at the Children’s Beach boat landing, was studied as well.5
(a)
(b)
Figure 2: Children’s Beach Pump Station – Alternative Setup
(a) Discharge Piping, (b) New Outlet Chamber for Pumped Flow
5 Because the new outlet chamber could be designed to work with a much shorter discharge pipe it was felt that pump
startup conditions might be improved.
Inlet side of Pump Station Not Shown
Inlet side of Pump Station Not Shown
CFD Model Review
Page 12
4.2 Study Scenarios
The Blue Hill Hydraulics’ CFD models were used to analyze the operational scenarios listed in Tables
3(a) and 3(b).
Table 3(a): Blue Hill Hydraulics’ Study Scenarios with Existing Outlet Chamber
Scenario No. Calculation of…
1 Water Surface Elevations in Manhole as a Function of Tailwater Level (5 sec. startup)
2 Water Surface Elevations in Manhole as a Function of Startup Time (5-90 sec. startup)
3 Water Surface Elevations in Manhole as a Function of Pumping Rate (5 sec. startup)
Table 3(b): Blue Hill Hydraulics’ Study Scenarios with New Outlet Chamber
Scenario No. Calculation of…
1 Water Surface Elevations in Manhole as a Function of Tailwater Level (5 sec. startup)
For each study scenario, the amount of surcharging in the manhole that provides access to the outlet
chamber was calculated and used as a “measure of success.” To be clearer, one could envision
removing the manhole cover and replacing it with a vertical section of pipe with a diameter the same as
that of the manhole. For a given model condition water surface elevations in the vertical section of pipe
were recorded, and at the conclusion of the tests a comparison of calculated water surface elevations was
made. Study scenarios that resulted in surcharging of the pump station were viewed to be failures and
those that did not produce any surcharging were viewed to be successes. This measure of success was
used because it was not possible to equate electrical demand with flow conditions in the pump station
A Note Concerning Datums: Two different datums are referenced on drawings and in correspondence
pertaining to this project. These are the “Nantucket Tidal Datum (NTD)” and the “Nantucket Half Tide
Level of 1934 (NHTL).” The NHTL 1934 was used in the Blue Hill Hydraulics study. Figure 3 shows
a cross-section of the pump station with NHTL elevations noted, and Table 4 shows the relationship
between different reference elevations based on the NTD and the NHTL. It is recommended that this
information be included in the final set of project drawings to prevent any confusion in the future.
CFD Model Review
Page 13
Figure 3: Cross Section and Datum Level
Table 4: Reference Elevations
Description
Nantucket
Tidal Datum
(ft)
Nantucket Half
Tide Level of
1934 (ft)
Highest Observed Water Level (02/07/1978) 6.37 3.96
Mean Higher High Water (MHHW) 3.60 1.19
Mean High Water (MHW) 3.26 0.85
Mean Tide Level (MTL) 1.73 -0.68
Mean Low Water (MLW) 0.20 -2.21
Mean Lower Low Water (MLLW) 0.00 -2.41
Lowest Observed Water Level (02/21/1981) -1.90 -4.21
18 October 2009 Storm (high tide 12:44pm – heaviest rain between
10:00am and 2:00pm – heavy rain reported at 12:53pm – New Moon)
3.9 1.49
48 inch Outlet Pipe Invert Level 0.97 -1.44
48 inch Outlet Pipe Soffit Level 4.97 2.56
CFD Model Review
Page 14
4.3 Results
4.3.1 Existing Outlet Chamber - Water Surface Elevations in Manhole as a Function of
Tailwater Level (5 Sec. Startups)
Results from six different calculations with varying tailwater levels were used to investigate changes in
pump station performance related to tidal water surface elevations in the harbor. As noted in Figure 4,
water surface elevations at the downstream end of the model were varied between 0.0 ft NHTL and 3.0
ft NHTL (note: 1.5 ft NHTL corresponds to the predicted tidal elevation for 18 October 2009 when the
pumps failed), and simulations of pump startup similar to those carried out by AECOM were performed.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0 102030405060708090100110120
time (sec)WSE (ft) NHTL0.00
1.00
1.50
2.00
2.50
3.00
Figure 4: Water Surface Elevations in Manhole during Startup
(downstream water surface elevations identified in legend, 5.0 sec. startup time)
The red line on Figure 4 identifies the street level elevation. Surcharging is predicted when the water
surface rises above this level and satisfactory operation of the pump station is assumed to take place
when the water surface in the pump station is below this level.
The results for simulations where the downstream water surface elevation is greater than or equal to 2.0
ft predict surcharging of the system, and the results for simulations where the downstream water surface
elevation is less than or equal to 1.5 ft are associated with conditions where satisfactory operation of the
CFD Model Review
Page 15
pump station could be expected. These results support the notion that satisfactory pump performance is
related to the water surface elevation in the harbor (i.e., transient pressures within the pump station vary
significantly when the tide is high, but they vary very little when the tide is low). These results also
agree reasonably well with the fact that a failure of the system occurred on 18 October 2009 when the
tide was high.6
4.3.2 Existing Outlet Chamber - Water Surface Elevations in Manhole as a Function of Pump
Startup Time (5 - 90 Sec.)
Four simulations were carried out to investigate whether or not variable speed pumps could be used to
eliminate surcharging of the pump station. In these simulations the downstream water surface elevation
was set at 2.5 ft NHTL (about 1.0 ft higher than the predicted water surface elevation in the harbor on 18
October 2009 to be conservative) and startup times were varied from 5 seconds to 90 seconds (note: the
startup of the pumps that are in place today was model by AECOM to take place over a 5 second period
and the same approach was used in these calculations).
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0 102030405060708090100110120
time (sec)WSE (ft) NHTL05 sec
30 sec
60 sec
90 sec
Figure 5: Water Surface Elevations in Manhole during Startup
(pump startup time identified in legend, downstream w.s.el. +2.5ft NHTL)
6 NOTE: These results do not predict surcharging when the downstream water surface elevation is equal to 1.5 ft NHTL.
However, we do not know what the exact operational conditions that led to the pumps shutdown on 18 October 2009
were and it is observed that the model results are sensitive to small changes in water surface elevation and boundary
conditions when the water surface elevation is between 1.5 ft NHTL and 2.0 ft NHTL, so in a general sense the model
results are consistent with observations.
CFD Model Review
Page 16
According to the results of these calculations, surcharging can be eliminated if the startup time of the
pumps is increased to 90 seconds. This allows time for the pumps to overcome the inertia of the water
in the discharge pipe before pumping rates are maximized and a change such as this would require that
modifications to the existing pumps be made to enable them to be operated at variable speeds.
4.3.3 Existing Outlet Chamber - Water Surface Elevations in Manhole as a Function of Pumping
Rate (5 Sec. Startups)
The results of five different simulations were used to determine if surcharging could be eliminated by a
reduction in pumping rates. In these calculations, pumping rates were reduced to 50% of their current
value and then they were further reduced in increments of 10% to a minimum pumping rate of about
11.2 cfs. As before, the downstream water surface elevation was conservatively set at 2.5 ft NHTL. As
shown in Figure 6, surcharging can be eliminated if pumping rates are reduced to less than 30% of their
current value.7
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0 102030405060708090100110120
time (sec)WSE (ft) NHTL100%
50%
40%
30%
20%
Figure 6: Water Surface Elevations in Manhole during Startup
(pumping rates identified in legend, downstream w.s.el. +2.5ft NHTL)
7 NOTE: The implication of this change to the upstream drainage basin was not studied.
100% Pumping Rate
equals 55.7 cfs
CFD Model Review
Page 17
4.3.4 New Outlet Chamber - Water Surface Elevations in Manhole as a Function of Tailwater
Level (5 Sec. Startups)
Similar to the results reported in Section 4.3.1, simulations of surcharging in a new outlet chamber that
would discharge to Children’s Beach were carried out. It was hoped that this alternative arrangement
would work better than the current one because it would rely on a shorter discharge pipe. However, we
calculated results that were virtually the same as those shown in Figure 4. Because this alternative
would require the construction of a new outlet chamber and discharge pipe in addition to making
changes to the pumps - the results of this analysis were not reported and this alternative was removed
from further consideration at this time.
5. Conclusions and Recommendations
The results of Blue Hill Hydraulics’ analysis agree with AECOM’s conclusion that successful startup of
the Children’s Beach pump station is related to tidal water surface elevations at the point of discharge
(i.e., tide levels in the yacht basin). When the tide is high, seawater fills the pump station’s discharge
pipe and when the pumps are activated they cannot set the seawater into motion without pressurizing the
outlet chamber. Because the pumps are designed to function at a constant speed, they draw increasing
amounts of electricity as the outlet chamber pressurizes. If the pumps cannot overcome the inertia of the
seawater in the discharge pipe before excessive electrical demands are made, then the pumps shutdown.
This could explain why the pump station did not start up properly on 18 October 2009 when heavy rains
fell during a high tide.8
As a result of this analysis, it is recommend that:
(1) Variable Frequency Drives (VFD) be used to control the startup and shutdown of
the pumps installed at Children’s Beach. The addition of these drives would not
require replacement of the existing pumps and the addition of the new electronic
control systems would allow for fine tuning of the operating system in the field
(e.g., adjusting startup speeds, reducing pumping rates, and limiting electrical
demands).9
and
(2) Vents be used to reduce the variation of pumping pressures. Some guidance
regarding vent size and design was provided by AECOM, but further analysis
should be carried out before any final decisions are made.
Finally, with regards to the discharge of water from the pump station, we would recommend that a
mixing analysis be carried out to determine the likelihood that changes made to the end-of-pipe would
be capable of reducing or eliminating the discoloring of boats in the harbor area. While not discussed
explicitly in this report, the discharge location could be moved. This change would, however, require
realignment of the pump station’s discharge pipe and would still require modification of the pumping
8 Heavy rains were reported between 10:00 am and 2:00 pm, and a high tide was predicted to occur at 12:44pm. 9 A cost estimate provided by Mr. Dan Weaver (Adams Equipment, Inc.) is provided in Appendix One of this report.
CFD Model Review
Page 18
system. Further consideration of these changes was outside the scope of this review, and it is not
recommended that the discharge location be moved at this time.
References
AECOM (2010), “Children’s Beach Stormwater Pump CFD Analysis,” PowerPoint presentation slides,
dated 8 March 2010.
Mahmutoglu, Serkan (2010), AECOM model input – prepin.inp files.
Mahmutoglu, Serkan (2010), AECOM model output – flsgrf.dat files.
Parece, Thomas (2010), Email correspondence containing requested hydrologic data from 15 August
through 29 October 2010.
Parece, Thomas (2010), “Nantucket, MA – Children’s Beach Stormwater System Evaluation and
Proposed Modifications,” AECOM memo to Mr. Jeffrey Willett, Director, Town of Nantucket,
Department of Public Works, dated 8 March 2010.
Richardson, John (2010), Blue Hill Hydraulics’ data request submitted to Mr. Jeffrey Willett for
approval, dated 25 August 2010.
CFD Model Review
Page 19
Appendix One: Cost Estimate for Installation of Variable Frequency Drives
This cost estimate was provided by Mr. Dan Weaver (Adams Equipment, Inc., phone 207.985.3828,
dan@adamsequipmentinc.com) in an email written to the author of this report on (3 April 2011).
Scope: Mount two VFDs into existing enclosure, wire, test, re-program and wire existing
pump controller for variable speed level control. Mount cooling fans and cut rain tight
louvers into enclosure sides for ventilation.
Assumptions: (1) One trip/two days on site, and
(2) Both VFDs can be mounted into the existing
enclosure (this assumption needs to be
confirmed with field measurements)
Cost: Two Yaskawa or Allen Bradley 40 HP VFD units $11,770.00
Labor $1,800.00
Materials $800.00
Travel and Expenses $500.00
Total (VFDs mounted in existing enclosure) $14,870.00
Total (VFDs mounted in additional enclosure) $15,670.00
NOTE: The cost estimate above does not include any permits, taxes, or additional
requirements not specifically noted above.